I turn AI experiments into reliable, production-ready systems — LLM agents, RAG pipelines, and automation workflows that actually ship.
Free · 30 minutes · No commitment
Specialized AI engineering for startups and scale-ups. I take projects from concept to production — fast, reliable, and built to last.
Production-ready LLM systems that go beyond demos — reliable pipelines, evaluated outputs, and cost-optimized deployments.
End-to-end workflow automation powered by AI — from document processing to multi-step agent pipelines with n8n at the core.
Turn your AI ideas into a concrete roadmap. From use-case discovery to architecture design and PoC delivery.
Recent AI projects — from concept to production.
Context-aware conversational AI agent delivered over WhatsApp. Built with LangChain for agent orchestration, a custom MCP server for structured tool calling, and LightRAG for knowledge graph-based long-term memory — enabling the assistant to retain context across sessions and reason over a connected knowledge base.
Open-source MCP server hub that dynamically loads 10,000+ tools at runtime via shapeshift(). Published on PyPI and npm, listed in the official MCP Registry. Features dynamic credential management, lean-mounting, and Smithers registry integration for zero-config tool discovery.
Production automation workflows built on n8n, orchestrating LLM agents, RAG pipelines, and third-party APIs to replace manual business processes. Covers document processing, lead qualification, data enrichment, and notification routing — all with error handling and human-in-the-loop escalation.
1st place at ElevenLabs Hackathon (n8n automation category), 1st place at NEAR AI Hackathon (AI agent category), and 1st place at AI Hardcore Hackathon (LLM cost & latency optimisation). Nine podium finishes total across AI-focused hackathons in 2025–2026.
A straightforward process built around shipping — not slides.
30 min · free
We talk through your current stack, what you've tried, and what success looks like for your project. No sales pitch — just an honest look at whether I'm the right fit.
2–3 days
I put together a clear scope: architecture, milestones, timeline, and deliverables. You know exactly what you're getting before anything starts.
Weekly updates
Production-first engineering from day one. Weekly check-ins, shared repo, and a running system you can test throughout — not just at the end.
Documentation included
Full documentation, knowledge transfer, and optional ongoing support. You own the system and can run it independently from day one.